The Use of Neutron Generators for the Detection of Hidden Explosives and Illicit Materials Giancarlo Nebbia Istituto Nazionale di Fisica Nucleare and Dip. di Fisica, Universita di Padova, Italy IAEA HQ 13 June 25
The detection of hidden explosives: a shift in the political agenda after 9/11? a) Detection of landmines buried by armies and/or irregular groups during recent conflicts b) Detection of unexploded ordnance from recent as well as old conflicts (i.e. World War II) c) Detection of explosive devices in the protection of sensible sites d) Detection of illicit trafficking of fissile and other threat materials (WMD) in the passengers and freight transportation system
Different tools for different needs For manual humanitarian demining one needs low-cost, handheld, operatorfriendly equipment to substitute/compliment the use of metal detectors. For military demining more sophisticated and expensive tools could be implemented.
Detection of explosives in the trade system : countermeasures against illicit trafficking The size of the container industry is enormous : in FY 22 the world s total movement in containers amounted to about 72 M TEU ( Twenty-foot Equivalent Unit) that are transported by ships and deposited inside the harbours Customs areas. Thereisanincreasingrisk that sizeable amounts of threat materials, including explosives, be hidden in cargo and transported by means of the standard commercial network.
Present inspection systems at ports are based on x-ray or γ-ray radiography. Courtesy of SAIC, San Diego, CA, USA Although the pictures are rather detailed and 3D imaging is possible, the number of suspect unidentified areas is still high.
Chemical composition of different materials (Courtesy of A.Buffler)
Neutron induced reactions Thermal neutron capture Inelastic scattering Backscattering
γ-ray transitions of relevant elements (MeV) Element capture inelastic H 2.2 none C 4.42 N 1.82 1.6-2.3-5.1 O 3.8 6.1
TNIS : Tagged Neutrons Inspection System based on the Associated Particle Technique In the d + t reaction a neutron with energy of 14 MeV and an alpha particle with energy of 3.5 MeV are emitted back-toback in the COM.
Why the associated particle technique? Gamma ray spectrum of a graphite sample without neutron tagging with the associated particle Gamma ray spectrum of the same graphite sample with neutron tagging by the associated particle Counts 15 Counts 4 3 1 2 5 1 2 3 4 5 6 7 Energy (MeV) 2 3 4 5 6 7 Energy (MeV)
VdG beamline with γ-ray detector array and soil box
The Associated Particle Detector (Phase I)
TDC Counts 9 8 7 6 5 4 I II III 3 2 1 12 122 124 126 128 Channel 13 132 134 Alpha-Gamma time-of-flight and gamma energy signals are recorded and used to recognize the elemental composition of a well defined irradiated area. Different elements are detected in different volume cells ( voxels ) using the tagged neutron beams.
The Associated Particle Detector (Phase II) YAP:Ce (YAlO 3 Cerium loaded) Φ = 4 mm, t =.5 mm coated with a layer of 1 mg/cm 2 of metallic silver Stainless steel CF63 flange PMT Hamamatsu R145 YAP:Ce UV-extended sapphire window Φ = 48 mm, t = 3 mm This detector is fully compatible with installation in a sealed neutron generator
Performances of the YAP(Ce) crystal in lab tests Energy resolution of YAP(Ce) for Eα = 5.4 MeV and Eγ = 59 KeV Timing resolution of two YAP(Ce) detectors for two coincident 511 KeV γ -rays
The experimental setup at the Institute Ruder Boskovic in Zagreb (Croatia) Beamlines dedicated to inspection of suitcases and containers for airport and harbour security
Irradiation of a 1x1x1 cm. graphite sample hidden inside the suitcase Right : alpha-gamma timing spectrum Left : gamma energy gated on the graphite in the timing spectrum
Inserting two different landmines (PROM-1 = 4 gr. of TNT and TMA-1A = 5 Kg. of TNT). Effect of the explosive on the gamma energy spectrum (gated on the alpha-gamma timing ).
A sealed neutron generator with the associated particle detector
First results with the sealed tube Counts 15 Time distribution (a) 1 Counts 15 (a) 1 5 5 1 2 3 4 5 6 7 8 9 Time (ns) 4 6 8 1 Time (ch) Counts 6 4 NaI(Tl) spectrum (b) Counts 6 (b) 4 2 2 1 2 3 4 5 6 Energy (MeV) 1 2 3 4 5 6 Energy (MeV) Counts 4 YAP:Ce spectrum (c) Counts 4 (c) 2 2 1 2 3 4 5 Energy (ch) 1 2 3 4 5 Energy (ch) Graphite sample irradiation spectra obtained from an NaI(Tl) detector without nanosecond coincidence with the α-particle Graphite sample irradiation spectra obtained from an NaI(Tl) detector with a 3 nanosecond coincidence with the α-particle
Layout of an inspection station Airport Luggage Sea Container There is need for 3D tagging capability!
The Associated Particle Detector (Phase III) YAP:Ce detector (Φ = 4 mm, t =.5 mm) read by 3 PMT's Hamamatsu R4141 (Φ = 13.5 mm)
Study of position sensitivity Φ PMT < Φ YAP:Ce => the light collection efficiency depends on the relative position of the alpha particles hitting the detector surface with respect to the PMT center (d source-pmt ) Counts Counts 22 2 18 16 14 12 1 8 6 4 2 X = 524 ch FWHM = 215 ch X =1253 ch FWHM = 144 ch X = 962 ch FWHM = 182 ch 2 4 6 8 1 12 14 Energy Channels (ch) 1 Assumption: Gaussian pulse height distributions dependent only on d source-pmt Parameterization of the amplitude and width values as a function of d source-pmt.8.6.4.2 Amplitude Sigma 2 4 6 8 1 12 14 16 18 d source-pmt center
Reconstruction algorithm P i calc : (A i1 calc, A i2 calc, A i3 calc ) (σ i1 calc, σ i2 calc, σ i3 calc ) P exp : (A 1 exp, A 2 exp, A 3 exp ) PMT 2 Y PMT 3 d d 3 2 X P(x,y) d 1 PMT 1 f ( x, y) = 3 i= 1 A => σ exp i (x rec,y rec ) calc i calc Ai ( x, ( x, y) y) Looking for the coordinates that minimize the function f(x,y) 2 15 1 5-6 -4-2 2 4 y (mm) 4 P (x,y) Entries 2338 Mean x.9359 Mean y -1.997 RMS x.8258 RMS y.7255 2 x (mm) -4-2
Test of the reconstruction algorithm y (mm) 8 6 4 PMT2 PMT3 2-2 -4 Φ holes = 1 mm -6-8 PMT1 Reconstructed position -1-8 -6-4 -2 2 4 6 8 x (mm)
The Associated Particle Detector (Phase IV) YAP:Ce detector (Φ = 4 mm, h =.5 mm) read by a 2x2 multi-anode PMT Hamamatsu R59U-- M4 (18x18 mm 2 )
Test of the reconstruction algorithm 14 12 1 y (mm) 1 5 PMT4 A C PMT1 B D 8 6 4 2 1 8 x (mm) 6 4 2-2 1 8 6 y (mm) 4 2-2 -5 PMT3 PMT2 Position A RMS x.7 RMS y.7-1 B C.5.9.7.9-1 -5 5 1 x (mm) D 1. 1.
Test with a cross collimator Collimator with two crossed 7 1 mm 2 slits placed in front of the YAP:Ce detector 8 6 4 2-6 -4-2 2 4 y (mm) 6 6 4 2 x (mm) -2-4 -6
Test with 2 graphite samples Neutron beam y A B x Sample A B x A (cm) -4 4 y A (cm) 4-4 x rec (mm) -2. 5.2 y rec (mm) 4. -3.3 y (mm) 15 1 6 5 5-5 -1-15 -15-1 -5 5 1 15 x (mm) 4 3 2 1 7 6 5 4 3 2 1 15 1 y (mm) 5-5 -1-5 -15-1 -15 5 1 15 x (mm)
Results from the Center of Gravity method Pros: Use of a single, large YAP(Ce) crystal Position resolution of the order of 2 mm (may improve to 1 mm) Cons: Analysis of amplitude signals -> position resolution depends on precise spectroscopy of the alpha signal Data acquisition rate of spectroscopic signal could be rather high (up to 1 7 counts/second) Is it possible to achieve a suitable position resolution by simple threshold discrimination on the fast PMT signals?
The Associated Particle Detector (Phase V) 64-elements 6x6 mm YAP(Ce) crystal matrix 64-pixels 6x6 mm H85 Hamamatsu PMT
The H85 PMT was coupled to the matrix YAP(Ce) crystals by a 4 mm thick quartz window. The system was irradiated with an alpha source through a 2 mm pinhole collimator
The four crystals indicated by the red square have been irradiated simultaneously through a square collimator. The red arrows show the position of the direct alpha hit signal in the amplitude spectrum for each crystal.
Setting a low threshold on signal #36 (pink area) results in cutting the direct alpha hit signal in #27 which is the farthest away. Setting a high threshold on signal #36 results in cutting the direct alpha hit signals in #28 and #35 which are adjacent to #36. In the response amplitude spectra of all crystals (pixels) the signal from direct alpha hit and from hits on adjacent crystals are totally decoupled. In this configuration one can use the threshold to determine position
For applications that require more performing position resolution new solutions are needed for the alpha tracking system What next?
The microstructure needle-pad detector Pre Amp Copper -V drift Pre Amp G-1 Needle: Anode +HV (2V) +HV Copper:Cathode
2D detector Active Area Detector holder Electronics 12x12 mm 2
Working principle X-ray tube
Position read-out Photon X1 X2 P x = X1 + X2 Y1 Y2 P Y = Y1 + Y2 X1 +++ + + +++ ++ ++ Y2 Y1 +++ +++ + + ++ X2
2D matrix detector X2 Y2 Y1 X1
Image of a half razor blade
The position resolution is of the order of 15 µm FWHM
The Needle-Pad detector works in any multiplication gas ( the residual gas inside a sealed neutron generator would be a good option) but the pressure needed to provide the electron-ion track is too high for applications inside a neutron generator. One needs a first stage electron converter and amplifier.
Widely used electron multipliers Microchannelplate detector Main problem: good vacuum required for operation (1-6 torr) Microsphere detector These electron multipliers have an average gain of 1 5 1 7. Biased at less than nominal voltage they can yield amplifications largely sufficient to work as a first stage converter and it has been demonstrated that they can work in much worse vacuum conditions (around 1-3 torr)
Conclusions: The use of the associated particle technique to tag 14 MeV neutrons for inspection of large items improves the quality of the γ-ray spectra largely reducing the background (experimental results on sealed generators show a factor 5 improvement on S/N ) Inspection of large items with miscellaneous loads (like a suitcase in an airport or a maritime container) requires the identification of a suitable size voxel to be irradiated The use of YAP(Ce) scintillators as alpha particle detectors for the tagging system is fully compatible with a sealed neutron generator and allows some pixelation, but other options are being considered to obtain better performances It is possible to reach a voxel size of few cm (2-3) in 2D in a close-in inspection system for suitcases and of about 3x3x3 cm 3 in any location inside a container using different position sensitive PMTs Reaching sub-millimeter position resolution would allow to apply APT to other inspection methods (i.e. fast neutron radiography) that require real imaging performances
First test on the detection of fissile material
TOF spectra for Pb and DU
Time of Flight spectra with γ γ coincidences Three different samples with the same weight have been irradiated for few minutes. Alpha-gamma-gamma triple coincidences have been recorded. In the Iron and Lead case one sees a very small of coincidences due to the detection of gammas in cascade. In the DU case one can notice the increase of coincidences due to the high multiplicity of fission fragment gamma decay.